The ventilatory and cerebrovascular responses to hypoxia provide the major short-term defenses against inadequate oxygen supply to the brain when that supply is threatened in situations as diverse as acute and chronic lung disease, cardiovascular disease and ascent to high altitude. These responses are not independent variables but interact substantially with each other. Four hypotheses regarding the way in which brain blood flow and/or brain tissue hypoxia may interact with respiratory control are proposed. (1) It is proposed that surges of brain blood flow (BBF) during REM sleep decrease respiratory neuronal output by means of reduction of PCO2 at the central chemoreceptor. This will be tested by (a) continuous simultaneous monitoring of BBF, brain venous pH, ventilation and respiratory muscle EMG; (b) mechanical restriction of BBF; and (c) antagonism of a proposed mediator of the BBF surge, dopamine. (2) It is proposed that during slow wave sleep the BBF response to hypoxia is such that the medulla and its chemoreceptors are preferentially perfused relative to the cortex, as is true in the awake state, but that this differential response is lost in REM sleep and with prolonged hypoxia. This will be tested by serial determinations of regional BBF in the various states using radioactive microspheres. (3) It is proposed that depression of ventilation due to brain hypoxia may be caused by elaboration of two major putative inhibitory neurotransmitters, endogenous opioids and Gamma aminobutyric acid. This will be tested by studies involving selective application of several antagonists and agonists of these agents as well as by measuring brain tissue levels of Gamma aminobutyric acid during brain hypoxia. (4) It is proposed that hypoxic stimulation of the carotid bodies increases the metabolism of ventral medullary respiratory nuclei sufficient to cause stimulation of the central chemoreceptors by their metabolic products, thereby establishing a positive feedback loop in respiratory control. This will be tested by measuring blood flow (uptake of iodoantipyrine) and metabolism (uptake of 2 deoxyglucose) in medullary respiratory nuclei during stimulation of the carotid sinus nerves and by measuring pH at the ventral surface of the medulla during stimulation of the carotid sinus and other sensory nerves.